The serine/threonine kinase Akt is one of the most frequently activated protein kinases in human cancers, and therefore represents an attractive potential target for therapeutic intervention. However, it is not completely understood how activation of Akt contributes to the genesis of cancer, and whether it is feasible to ablate or reduce Akt activity in order to cure cancer. Our long-term goals are: 1) To determine why Akt is so frequently activated in human cancers. 2) To elucidate the mechanisms by which Akt promotes cell cycle progression and susceptibility to oncogenic transformation both at the cellular and organism levels. 3) To determine if it is feasible to partially ablate Akt activity to inhibit cancer development without any other severe physiological consequences. For this purpose we have generated mice deficient for the three Akt isoforms (Akt1-3) individually as well as compound mutant mice with all possible combined deletions. We are analyzing the phenotypes of these mice and examining their susceptibility to tumorigenesis. We will determine how much of Akt activity could be reduced to inhibit tumorigenesis, and could be tolerated without severe physiological consequences, and which Akt isoform is a better target for cancer therapy. Analysis of cells derived from these mice shows that Akt is required for normal cellular proliferation and susceptibility to oncogenic transformation. We are delineating the mechanisms by which Akt exerts its effect on cell cycle progression and susceptibility to oncogenic transformation. These studies will eventually determine the most critical downstream effectors of Akt required for these processes. The first part of this grant application will address the role of Akt in cell cycle progression, G2 cell cycle checkpoint, and oncogenic transformation at the cellular level. The second part of this grant application will address these issues at the organism level and assess the susceptibility of various Akt knockout (KO) mice to the development of various neoplasia. [unreadable] [unreadable] [unreadable]